摘要:
Embodiments of the invention are directed to providing a method for selecting a link for transmitting a data packet, from links of a Multi-Link Point-to-Point Protocol (MLPPP) bundle, by compiling a list of links having a minimum queue depth and selecting the link in a round robin manner from the list. Some embodiments of the invention further provide for a flag to indicate if the selected link has been assigned to a transmitter so that an appropriate link will be selected even if link queue depth status is not current.
摘要:
Various exemplary embodiments relate to a method and related network node including one or more of the following: maintaining a line card activity number at each line card in the node; receiving a first packet belonging to a flow; storing, at each line card, a flow record in a line card context identifier storage, wherein the flow record contains a flow activity number being initially set equal to a current value of the line card activity number; detecting, by the second line card, a change in designation to active and changing the value of the line card activity number maintained by the second line card; receiving a second packet belonging to the flow; and transmitting, to the decompressor node, an update packet including the flow context identifier and the flow identifying information, upon determining that the flow activity number is not equal to the updated line card activity number.
摘要:
A method and apparatus for handling, maintaining, and controlling network synchronization information emanating from a plurality of line card circuits is described. The technique described may be applied to a redundant pair of line card circuits, where one line card circuit is active, while the other is inactive. Line card activity latches are managed by means of hardware logic that may be configured at the time of line card commissioning. The activity latches are coupled to a logic element. An incoming clock signal is applied to the logic element. If an activity latch indicates that a line card circuit is active, the logic element provides the incoming clock signal as an outgoing clock signal to a control card circuit. If the activity latch indicates that the line card circuit is inactive, the logic element blocks the incoming clock signal from being passed and provides a static output level as the outgoing clock signal to the control card circuit. The control card circuit is provided with circuitry to receive the outgoing clock signals from multiple line card circuits. The circuitry is sensitive to whether or not the line card circuits are configured for redundant operation. One or more of these clock signals are then selected and used for network synchronization.
摘要:
The invention is directed to a method and system for efficiently distributing Inverse Multiplexing over ATM (IMA) filler cells on IMA logical links having inactive or unusable IMA sub-links. Inactive IMA sub-links are identified during round-robin distribution of ATM data cells to active IMA sub-links. Predefined IMA filler cells are transmitted simultaneously on the identified inactive IMA sub-links when an ATM data cell is transmitted on an active IMA sub-link, thereby reducing the delay between transmitting ATM data cells.
摘要:
Various exemplary embodiments relate to a method and related network node including one or more of the following: determining that a first search value is associated with a first range field; determining a first bitmap associated with the first search value, wherein the first bitmap indicates at least one range encompassing the first search value; generating a search key based on the first bitmap; and accessing the ternary content addressable memory based on the search key.
摘要:
Various exemplary embodiments relate to a method and related network node having a playout buffer including one or more of the following: receiving, at the network device, a first packet belonging to a first flow, the first packet including a first sequence number (SN); receiving, at the network device, a second packet belonging to the first flow, the second packet including a second SN; determining that the second SN is not in sequence with the first SN; waiting to receive, at the network device, a third packet belonging to the first flow, the third packet including a third SN; and determining that a jump in SNs has occurred for the first flow between the first packet and the second packet based on determining that the third SN is in sequence with the second SN.
摘要:
Various exemplary embodiments relate to a method and related network node including one or more of the following: receiving a packet belonging to a flow, the packet including a marking used to identify a quality of service (QoS) required for the packet; performing buffer management to determine whether a current occupancy level of a buffer in the network node is greater than a threshold level; when the current occupancy level of the buffer does not exceed the threshold level, performing header compression on the packet in the network node; after performing header compression, performing traffic scheduling to queue the packet in a queue selected from a plurality of queues based on the marking in the packet and to output the packet from the selected queue; and forwarding the packet towards a destination, whereby buffer management, header compression, and traffic scheduling are all performed in a single network node.
摘要:
Various exemplary embodiments relate to a method and related network node having a playout buffer including one or more of the following: receiving a first packet, a second packet, a first set of at least one subsequent packet, wherein each packet includes a sequence number (SN); determining that the second packet is not in sequence with the first packet by determining that the SN of the second packet is not equal to the SN of the first packet plus an expected increment value; determining whether the second packet represents a jump in SNs by determining whether the SN of a first subsequent packet is equal to the SN of the second packet plus the expected increment value; and when the second packet represents a jump in SNs, gradually normalizing the playout buffer upon receipt of each subsequent packet.
摘要:
Various exemplary embodiments relate to a method and related network node including one or more of the following: receiving a packet belonging to a flow, the packet including a marking used to identify a quality of service (QoS) required for the packet; performing buffer management to determine whether a current occupancy level of a buffer in the network node is greater than a threshold level; when the current occupancy level of the buffer does not exceed the threshold level, performing header compression on the packet in the network node; after performing header compression, performing traffic scheduling to queue the packet in a queue selected from a plurality of queues based on the marking in the packet and to output the packet from the selected queue; and forwarding the packet towards a destination, whereby buffer management, header compression, and traffic scheduling are all performed in a single network node.
摘要:
Various exemplary embodiments relate to a method and related network node including one or more of the following: establishing a plurality of flows; assigning a context identifier to each flow, wherein a length of each context identifier for a first subset of the flows is a first number of bits and a length of each context identifier for a second subset of the flows is a second number of bits greater than the first number of bits; receiving a packet associated with a new flow, wherein a value of a context identifier assigned to the new flow must fit within the first number of bits; determining that the context identifiers for the plurality of flows have occupied all values within the first number of bits; assigning a context identifier of an existing flow to the new flow; and assigning a new context identifier to the existing flow.